The Microtubule Regulator NEK7 Coordinates the Wiring of Cortical Parvalbumin Interneurons  Antonio Jesús Hinojosa, Rubén Deogracias, Beatriz Rico  Cell Reports  Volume 24, Issue 5, Pages 1231-1242 (July 2018) DOI: 10.1016/j.celrep.2018.06.115 Copyright © 2018 Terms and Conditions

Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions

Figure 1 Nek7 Transcripts Are Upregulated during GABAergic Interneuron Wiring and Expressed Mostly in PV+ Cells (A) Schematic of genetic screening. Cortices from different GFP+ reporter lines were dissected, and cells were dissociated and FACS sorted and their RNA hybridized to microarrays. (B) Bioinformatic comparisons of neuronal populations. (C) Heatmap showing the expression levels of the 20 first ranked genes. Nek7 is highlighted in red. (D) Confocal images showing in situ hybridization for Nek7 (red) and immunohistochemistry for GFP (green) and SST+ (cyan) in the somatosensory cortex of P10 Lhx6Cre;RCE mice. GFP+ and SST+ are somatostatin cells, GFP+ and SST− are putative PV+ interneurons. (E) Confocal images showing in situ hybridization for Nek7 (red) and immunohistochemistry for PV+ cells (green) in the somatosensory cortex of P30 wild-type mice. In (D) and (E), filled arrowheads denote colocalization, and open arrowheads denote no colocalization. (F) Percentage of PV+ (gray) among all Nek7-expressing cortical cells at P10 and P30. (G) Percentage of Nek7-positive cells among all PV-expressing neurons at P10 and P30. IN, interneurons (red); Pyr, pyramidal cells (blue). Data are represented as mean ± SEM. Scale bar represents 50 μm. See also Figures S1–S4. Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions

Figure 2 Nek7 Knockdown Accelerates Microtubule Growth (A) Diagram of the Cre-dependent constructs expressing mCherry and shRNA. The plasmids were co-transfected with EB3-YFP, and mNek7 for the rescue, in primary cortical cultures from Nkx2-1Cre mice at 4 DIV, and axons were recorded at 7 DIV. (B–D) Confocal Z projection frames from control shRNA (B), Nek7 shRNA (C), and rescue with mNek7 (D) Nkx2-1Cre growth cones expressing mCherry (red, before the time-lapse) and EB3-YFP (gray, time-lapse). The path of EB3-YFP comets is tracked with a red line. Scale bar represents 2 μm. (E–G) Kymographs of EB3 comets in control (E), Nek7 shRNA (F), and rescue (G) growth cones showing their existence time as a function of distance. (H and I) Average speed (H) and speed distribution of EB3 comets (I) comparing control (n = 31 growth cones), Nek7-depleted cells (n = 35 growth cones), and mNek7 rescue cells (n = 35 growth cones) from three independent cultures. All growth cones with EB3 comets were quantified. One-way ANOVA (H) and χ2 test (I). ∗p < 0.05 and ∗∗p < 0.01. Data are represented as mean ± SEM (H) or total cell percentage (I). See also Figures S5–S7 and Videos S1, S2, and S3. Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions

Figure 3 Nek7 Depletion Impairs Axonal Growth Cone Dynamics In Vitro (A) Schematic of experimental design with the Cre-dependent plasmids transfected in Nkx2-1Cre primary cortical cultures at 4 DIV and axons recorded at 7 DIV. (B–D) Confocal Z projection frames from axons expressing control (B), Nek7 shRNA (C), and mNek7 rescue (D). The last frame in the sequence shows superimposed images of the frames t = 0 min (red), t = 15 min (green), t = 30 min (blue), and t = 45 min (white). Scale bar represents 5 μm. (E–G) Axonal speed (E), average of growth cone maximum turning angle (F), and its distribution (G) from cells expressing control shRNA (n = 64 growth cones), Nek7 shRNA (n = 59 growth cones), and mNek7 (n = 85 growth cones) from three or four independent cultures. All growth cones increasing in length were quantified. Kruskal-Wallis test, pairwise comparisons (E and F), and χ2 test (G). Data are represented as mean ± SEM (E and F) or total cell percentage (G). ∗∗p < 0.01 and ∗∗∗p < 0.001; n.s., not significant. See also Figures S6 and S7 and Videos S4, S5, and S6. Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions

Figure 4 Loss of NEK7 Alters Interneuron Morphology In Vitro (A) Schematic of experimental design. pDIO-shRNA was transfected in Nkx2-1Cre primary cortical cultures. (B–D) Confocal Z projections of Nkx2-1Cre interneurons from cortical cultures transfected with control shRNA (B), Nek7 shRNA (C), and mNek7 for rescue (D) expressing mCherry. The cells were automatically reconstructed and masked in black at 12 DIV. Scale bars represent 200 μm. (E–I) Average of total neurite length (E) and its distribution (F), Sholl analysis (G), total branching points (H), and branching points per unit length (I) of control shRNA (n = 47), Nek7 shRNA (n = 45), and rescue (n = 46) transfected neurons from three or four independent cultures. All imaged cells were quantified. Kruskal-Wallis test, pairwise comparisons (E, H, and I), χ2 test (F), and two-way ANOVA with Bonferroni correction (G). ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001. Data are represented as mean ± SEM. See also Figures S6 and S7. Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions

Figure 5 Nek7 Overexpression in Pyramids Does Not Alter Their Axonal Development (A and B) Confocal Z projection frames from NexCre growth cones expressing mCherry (red, before the time-lapse), EB3-YFP (gray, time-lapse), and mNek7 in the Nek7 condition. Control (A) and mNek7 (B). The path of EB3-YFP comets is tracked with a red line. (C and D) Confocal Z projection frames from NexCre growth cones. Control (C) and mNEK7 (D). The last frame in the sequence shows superimposed images of the frames t = 0 min (red), t = 15 min (green), t = 30 min (blue), and t = 45 min (white). (E and F) Confocal Z projections of NexCre interneurons expressing mCherry. Control (E) and mNek7 (F). The cells were automatically reconstructed and masked in black at 8 DIV. (G) Average speed of EB3 comets comparing control pyramids (A; n = 53 growth cones) and Nek7-expressing pyramids (B; n = 54 growth cones) from three independent cultures. (H and I) Average of growth cone maximum turning angle (H) and axonal speed (I) from control pyramids (C; n = 73 growth cones) and Nek7-expressing pyramids (D; n = 74 growth cones) from three independent cultures. (J and K) Neurite length (J) and branching points (K) comparing control pyramids (E; n = 44 cells) and Nek7-expressing pyramids (F; n = 38 cells) transfected neurons from three independent cultures. One-way ANOVA (G and I) and Mann-Whitney test (H, J, and K). n.s., not significant. Data are represented as mean ± SEM. Scale bars represent 2 μm (A and B), 5 μm (C and D), and 200 μm (E and F). Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions

Figure 6 Loss of NEK7 Causes Abnormal PV+ Neuronal Morphology In Vivo (A) Schematic of experimental design. Cre-dependent AAVs expressing shRNA and the fluorescent marker mCherry were injected in E15.5 Lhx6Cre mice in vivo. (B and C) Targeted mCherry+ interneurons (B) that were PV+ (C) were selected for the analysis. (D and E) Confocal Z projections of PV+ interneurons expressing control shRNA (D) and Nek7 shRNA (E) reported by mCherry. The cells were automatically reconstructed and masked in black from layer V of somatosensory cortex at P21. (F–I) Average of total neurite length (F), Sholl analysis (G), total branching points (H), and branching points per 100 μm of neurite (I) of control shRNA (n = 30 PV+ neurons, from 10 mice) and Nek7 shRNA (n = 27 PV+ neurons, from 8 mice) infected neurons. One-way ANOVA (F, H, and I) and two-way ANOVA with Bonferroni correction (G). ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001; n.s., not significant. Data are represented as mean ± SEM. Scale bars represent 200 μm (B) and 50 μm (C–E). See also Figure S6. Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions

Figure 7 Nek7 Knockdown Decreases PV+ Interneuron Outputs (A) Schematic of the experimental design. Cre-dependent virus expressing shRNA and the fluorescent marker mCherry were injected in E15.5 (left) and P3 (right) Lhx6Cre mice in vivo. Boutons co-stained with SYT2 and mCherry were quantified either inside PV+ arbors or onto NeuN+ somata. (B and C) Confocal Z projections and Imaris 3D reconstructions showing control shRNA (B) and Nek7 shRNA (C) infected PV+ cells expressing mCherry (red) in axons containing SYT2+ boutons (blue). (D) SYT2+ bouton density per unit area of neurite comparing control shRNA (n = 24 PV+ neurons from four mice) and Nek7 shRNA (n = 22 PV+ neurons from six mice). Student’s t test. (E–G) Confocal images and 3D Imaris reconstructions showing mCherry+ (red), SYT2+ (green) synaptic boutons from infected PV+ cells expressing control shRNA (E), Nek7 shRNA (F), and Nek7 shRNA with mNek7 for rescue (G) contacting pyramidal cells NeuN+ (blue). SYT2+ mCherry+ (filled arrowheads, yellow spheres), SYT2+ mCherry− (open arrowheads, green spheres). (H) Percentage of SYT2+ mCherry+ somatic boutons contacting the pyramidal cells normalized to the percentage of PV+ cells infected in the area, comparing shRNA (n = 182 PV+ neurons from seven mice), Nek7 shRNA (n = 200 pyramidal cells from six mice), and mNek7 (n = 109 pyramidal cells from four mice). Kruskal-Wallis test, pairwise comparisons. ∗p < 0.05 and ∗∗p < 0.01; n.s., not significant. Data are represented as mean ± SEM. Scale bars represent 2 μm (B and C) and 5 μm (E–G). See also Figure S6. Cell Reports 2018 24, 1231-1242DOI: (10.1016/j.celrep.2018.06.115) Copyright © 2018 Terms and Conditions